Blood pressure and mortality: an epidemiological survey with 10 years follow-up

The Copenhagen City Heart Study is a prospective ischaemic heart disease population study designed to evaluate incidence of, and risk factors for, ischaemic heart disease. A random population sample of approximately 20,000 men and women was invited to participate in a health survey, which was carried out in 1976-78. The participation rate was 74%. Systolic (SBP) and diastolic blood pressure (DBP) was measured with the London School of Hygiene sphygmomanometer after 5 minutes in the sitting position. Risk factors were assessed by a questionnaire and non-fasting plasma cholesterol was measured. Information about subsequent death and causes of death was obtained from the Danish Death Register. Follow-up was virtually complete over an observation time of 10 years. Analysis of the independent effect of SBP and DBP measured at entry on the 10 year total and cause-specific mortality was performed using the Cox regression model. Antihypertensive medication and/or diuretic therapy, physical activity during leisure time, economic and educational status, tobacco and alcohol consumption, diabetes mellitus, body mass index, plasma cholesterol levels, age and sex were entered as confounders. Total mortality was increased only in the higher quintiles of SBP. Concerning ischaemic heart disease mortality and cerebrovascular mortality, the risk increased in a graded manner with increasing quintile of SBP and DBP. With regard to cancer mortality, a U-shaped association was observed between quintile of SBP (and DBP) and death rate. With advancing age, the predictive power of SBP on total and cause-specific mortality changed, especially in males, as a pronounced U-shape of the association between BP and mortality appeared. The reasons for this are discussed. The relative risk in subjects receiving antihypertensive medication was 1.7 (CL 1.5-2.0) regarding total mortality, 2.0 (CL 1.5-2.7) regarding ischaemic heart disease mortality, 0.8 (CL 0.5-1.4) regarding cerebrovascular mortality, and 1.3 (CL 1.0-1.7) regarding cancer mortality. This finding is in agreement with clinical trials experiences, and may have an impact on management of high blood pressure.